Diagnostic and Statistical manual of Mental Disorders, Third Edition (DSM-III; American Psychiatric Association (APA), 1980) were empiri- cally-keyed criterion sets with extensive validation histories. The newest Manual, DSM-III-R (APA, 1987), is in general a mere "revision" of DSM-III: Few criterion sets were altered, and field testing has been deferred to the development of DSM-IV in the mid-1990's. The most common psychiatric entities, however, Psychoactive Substance Use Disorders (PSUDs), have been entirely re-keyed in DSM-III-R. The goal of this study is to test, in a fashion superior to the field testing of DSM-III, the reliability, specificity, sensitivity, arid forms of validity of the new diagnostic rules for PSUDs promulgated by DSM-III-R, and to suggest differential criterion weighting schemes. Four-hundred clinically diagnosable subjects and 135 "matched community control subjects" will be recruited in four Northeastern states, sampline a heterogeneous class of diagnostic targets ranging from "low-bottom" alcoholics at a V.A. hospital to affluent cocaine addicts in private treatment. Eighty clinical subjects will be diagnosed independently by twin raters, to determine the kappa reliability of DSM-III-R's rules. All subjects will be diagnosed psychiatrically bv formal instruments, and will be diagnosed for PSUDs according to the rules of DSFI-I@l. and DSM-III-R. Multiple severity ratings will be made, and independent measures of behavioral, social, occupational, family, legal and somatic complications will be taken. The relationship of these measures to specific diagnostic markers in DSM-III-R will be studied (concurrent, convergent and discriminant validity). Followed for two years by a consortium of seven treatment-contact sites, subjects will be re-assessed at intervals of 6, 12 and 24 months post-intake (predictive and construct validity). Especially problematic vs. promising markers and possible differential weighting schemes will be identified, as will the presence of specific, stable symptom constellations and their relationship to severity, clinical outcome, treatment response and concurrent psychopathology.